After the decomposition is obtained and a grouping is selected from analyzing the fssa object exploratory plots (see Functional Singular Spectrum Analysis: Functional Time Series Decomposition for more information) the user may choose to perform forecasting of the FTS. The fforecast function takes an argument of class fssa, the selected grouping as expressed in a list, and the forecast horizon and returns a list object that has the same length as the group argument where each entry is an fts object built from forecasting a specified group via FSSA recurrent or FSSA vector forecasting. We note that currently, forecasting is available only for FTS variables observed over a one-dimensional domain and forecasting FTS variables observed over a two-dimensional domain is under development. We also note that while it is possible to forecast different parts of the signal individually (such as only mean or only periodic natures), it is most common to forecast the whole assumed deterministic signal that was extracted by FSSA. The following code executes the forecasting process applied to the decompositions of the call center data (see Functional Singular Spectrum Analysis: Functional Time Series Objects and Visualization for more information) to predict three weeks into the future (21 days). As a reminder we give the following groupings for the call center dataset based on the plots of the decompositions.

## Call center data
gr_C <- list(1, 2:3, 4:5, 6:7, 1:7)
Q_C <- fforecast(U_C, gr_C, h = 21)

The following plots show the individual forecasted signals obtained from the decompositions as according to the specific groupings as well as the original FTS and the three weeks into the future forecast of the overall extracted signal. All forecasts were performed using the FSSA recurrent forecasting algorithm.

## Call center data
plot(Q_C[[1]], mains = "Call Center Mean Component Forecast",xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))
plot(Q_C[[2]], mains = "Call Center First Periodic Component Forecast",xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))
plot(Q_C[[3]], mains = "Call Center Second Periodic Component Forecast",xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))
plot(Q_C[[4]], mains = "Call Center Third Periodic Component Forecast",xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))
plot(Y_C, mains = c("Call Center Data Line Plot"),xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))
plot(Q_C[[5]], mains = "Call Center Extracted Signal Forecast",xlabels = "Time (6 minutes aggregated)",
ylabels = "Sqrt of Call Numbers",type="line",
xticklabels = list(c("00:00","06:00","12:00","18:00","24:00")),xticklocs =
 list(c(1,60,120,180,240)))

In the various plots, the user can investigate the nature and patterns present in different forecasts. In particular, the forecast of the extracted signal reflects that the FSSA-based forecasting technique does well in predicting periodic behaviors in the call center FTS.